• DocumentCode
    1591391
  • Title

    Day-to-day reproducibility and deviations from uniformity of two fractal measures of VPC clustering in patients with life threatening ventricular arrhythmias

  • Author

    Stein, KM ; Karagounis, La ; Anderson, J.L. ; Kligfield, P.

  • Author_Institution
    Cornell Med. Center, New York, NY, USA
  • fYear
    1992
  • Firstpage
    239
  • Lastpage
    241
  • Abstract
    The authors previously described a technique to measure clustering of ventricular ectopy (VE) by measuring a fractal dimension (D) for the first 200 ventricular premature contractions (VPCs) recorded (FirstD200) during ambulatory electrocardiography. Reduced FirstD200 reflects clustered VE and has been associated with poor prognosis but has been found to exhibit large day-to-day variability. A new algorithm was developed that measures the median D values of overlapping segments of 200 VPCs observed in a given day (MedD200). The authors compared the reproducibility of First D200 and MedD200 in 18 patients with a history of life-threatening ventricular arrhythmias who had two studies obtained in the drug-free state within 30 days. Each D value was paired with a value calculated by applying the same technique to a set of VPCs simulated by a uniform random distribution. The conclusion is that in this population, VPCs are nonuniformly distributed with respect to time and that the MedD200 technique provides a more reproducible measure of clustering than the FirstD200 technique
  • Keywords
    electrocardiography; fractals; medical signal processing; 30 d; algorithm; ambulatory electrocardiography; day-to-day reproducibility; drug-free state; fractal measures; life threatening ventricular arrhythmias; ventricular ectopy clustering; ventricular premature contractions; Cities and towns; Clustering algorithms; Electrocardiography; Fractals; History; Hospitals; Medical treatment; Pattern analysis; Reproducibility of results; Visual BASIC;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computers in Cardiology 1992, Proceedings of
  • Conference_Location
    Durham, NC
  • Print_ISBN
    0-8186-3552-5
  • Type

    conf

  • DOI
    10.1109/CIC.1992.269402
  • Filename
    269402